Skill Claude CodeCodex
Principal-engineer code review in two modes — pre-PR (local Standards + Spec axes via parallel sub-agents, nothing posted) and post-PR (11-dimension rubric, per-comment approval loop, posts one grouped GitHub Review). Invoked only by typing /code-review-multi-axis (Codex: $code-review-multi-axis).
Skill Claude CodeCodex
Compact a long conversation into a saved progress summary — completed work, key decisions, current state, pending tasks — so a fresh session can pick up without losing context. Fires only as a fallback when the agent has no native /compact command; check for one first. Use when the user says "compact this…
Skill Claude CodeCodex
Use when reporting status, progress, results, errors, or any information back to the user. Does not apply when writing content, documents, or artifacts for the user.
Skill Claude Code
Run the full UXD workflow end-to-end for an approved PRD — design interview → user flow → wireframes → design system → hi-fi screens → design review → developer handoff → approve. Reads approved PRDs from pm-os automatically.
Skill Claude Code
Interactive design brief interview. Always run before generating any screens. Extracts product context, user goals, visual style, constraints, and success criteria through a structured conversation.
Skill Claude Code
Multi-agent design review from 5 perspectives — UX Researcher, Accessibility Auditor, Engineer, Brand Guardian, and End User Voice. Feed it any generated screen file or design description.
Skill Claude Code
Engineering-only. Stress-tests an engineering artifact (eng spec, TDD plan, refactor proposal, architecture sketch, technical design, schema migration, API contract) against the project's CONTEXT.md. Reads CONTEXT.md read-only, interviews the user against existing terminology and decisions, surfaces conflicts, may add…
Skill Claude Code
Run a Playwright end-to-end audit of a web app using its PRDs as the test spec. Produces a diagnostic report showing what's working and what's not from a user's perspective.
Skill Claude CodeCodex
Convert an approved PRD into TDD + Engineering Spec + ADRs, run a 5-agent engineering panel review (Architect, DB Designer, Principal SWE, SDET, Performance Engineer), fix blockers, approve, and sync to ralph.
Skill Claude CodeCodex
Use when scaffolding a GitHub Pages site for a new (or existing) repo and you want a cohesive neo-brutalist design system. Drops in Jekyll layouts, design tokens, four-theme switcher (light/dark/cyberpunk/solarized), copy buttons, and a hero/nav/markdown-body system in one shot.
Skill Claude Code
Scaffold a new GitHub repo that mirrors an existing reference repo's general settings, branch-protection rules, security flags, and (optionally) its GitHub Pages site — porting the static HTML/CSS/JS docs and rebranding them. Use when the user says "create a new repo like X", "mirror settings from Y", "new repo same…
Skill Claude CodeCodex
Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.
Skill Claude CodeCodex
Use when the user wants a non-interactive Grill Me run written as one collapsible markdown document, asks for "Grill Me Auto", "auto grill", "all grill questions at once", or invokes /grill-me-auto, /Grill Me Auto, or /grill me auto.
Skill Claude CodeCodex
Reformat a status update so leadership remembers you positively. Turn raw notes into outcome-first, three-sentence updates with a clear ask. Use when the user wants to give an update to a manager, exec, or stakeholder, mentions a "status update", "leadership update", "exec update", or pastes raw notes about what…
Skill Claude CodeCodex
Research, compare, and select the best npm/pnpm/yarn packages before installing. Use when scaffolding a project, adding a dependency or devDependency, or choosing between competing libraries — including when the user says "add dependency" or "which library should I use".
Skill Claude CodeCodex
Full project status across Product (PM-OS), Engineering DOE (specs/ADRs), and Engineering Implementation (fixplan/ralph). Run this at the start of any session to orient yourself, or on demand.
Skill Claude Code
Approve a reviewed PRD — updates status, copies to approved inbox, and updates PRD-PIPELINE.md. Run only after /prd-review-panel is complete and you are satisfied with the output.
Skill Claude Code
Create a modern, AI-era PRD for features and initiatives. Guides through clarifying questions, generates draft, and offers multi-agent review.
Skill Claude CodeCodex
Run a 7-perspective multi-agent PRD review (Engineering, Design, Executive, Legal, UXR, Skeptic, Customer Voice) that catches gaps and surfaces conflicts before stakeholder review. Use when the user wants a PRD reviewed, asks for panel or multi-agent feedback, or runs /prd-review-panel.
Skill Claude CodeCodex
Turn a PRD into a multi-phase implementation plan using tracer-bullet vertical slices, saved as a local Markdown file in ./plans/. Use when the user wants to break a PRD into an implementation plan, or mentions "tracer bullets".
Skill Claude CodeCodex
Universal conciseness and precision directive for all LLM output. Prepend to any system prompt to eliminate verbosity, filler, and fluff from every response.
Skill Claude CodeCodex
Interactive QA session where user reports bugs or issues conversationally, and the agent files GitHub issues. Explores the codebase in the background for context and domain language. Use when user wants to report bugs, do QA, file issues conversationally, or mentions "QA session".
Skill Claude CodeCodex
Scans a codebase to build CONTEXT.md (or CONTEXT-MAP.md for multi-context repos) and seed ADRs for clearly-deliberate decisions. Autonomous by default; asks only when ambiguity blocks resolution. Use after cloning a repo, joining a project mid-flight, or when docs have drifted from code. Triggers include "scan repo…
Skill Claude CodeCodex
Create a detailed refactor plan with tiny commits via user interview, then file it as a GitHub issue. Use when user wants to plan a refactor, create a refactoring RFC, or break a refactor into safe incremental steps.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: